A population-based case-control study of prostate cancer in African American men is in progress in Genesee County, Michigan. The original objective of describing the age-specific reference values for PSA in a random sample of African Americans, 40-79 years of age, without clinical evidence or a history of prostate cancer, has been initiated through the completion of 833 in-home interviews, 382 PSA assays and 289 urological examinations. The target of completing 500 PSA assays (total PSA and ratio of free to total PSA) and urological examinations should be reached by March, 1998. During years 03, 04, and 05, we are projecting that 150-200 newly diagnosed prostate cancer cancers in African Americans residing in Genesee County will be participants in the study. For the random sample of African Americans, an upper 95% prediction limit will be constructed around the mean PSA values in 10-year age groups. These reference ranges are so designed to be compatible with 95% specificity. The analysis will also describe the 5th, 25th, 50th, 75th and 95th percentile for the log- normal distributions of PSA concentrations by 10-year age groups in the random controls and prostate cancer cases. The clinical utility of age- specific reference values in African Americans will be underscored by the distribution of false positives in the controls and the true positives in the cases. New directions in research utilizing the case-control design, will focus on the polymorphic variations in the androgen receptor (AR) gene, type II 5 alpha-reductase (SRD 5A2) gene and vitamin D receptor (VDR) gene in African Americans. Genetic factors may be contributing in fundamental ways to black-white differences in prostate cancer incidence, particularly in relation to androgen-responsive genes and "hormonal" nutrients that serve to modulate proliferation and differentiation. Genetic epidemiologic studies have been facilitated by the development of a resource consisting of a comprehensive data base of epidemiologic risk factors in random controls and population-based cases; frozen stormed serum, plasma and DNA in controls and cases; and frozen and paraffin blocks from the cases diagnosed in 1997 and in subsequent years.